Skip to content

Query Understanding

Stop using LGTM@Few as a metric (Better RAG)

I work with a few seed series a startups that are ramping out their retrieval augmented generation systems. I've noticed a lot of unclear thinking around what metrics to use and when to use them. I've seen a lot of people use "LGTM@Few" as a metric, and I think it's a terrible idea. I'm going to explain why and what you should use instead.

If you want to learn about my consulting practice check out my services page. If you're interested in working together please reach out to me via email

When giving advice to developers on improving their retrieval augmented generation, I usually say two things:

  1. Look at the Data
  2. Don't just look at the Data

Wise men speak in paradoxes because we are afraid of half-truths. This blog post will try to capture when to look at data and when to stop looking at data in the context of retrieval augmented generation.

I'll cover the different relevancy and ranking metrics, some stories to help you understand them, their trade-offs, and some general advice on how to think.

How to build a terrible RAG system

If you've seen any of my work, you know that the main message I have for anyone building a RAG system is to think of it primarily as a recommendation system. Today, I want to introduce the concept of inverted thinking to address how we should approach the challenge of creating an exceptional system.

What is inverted thinking?

Inversion is the practice of thinking through problems in reverse. It's the practice of “inverting” a problem - turning it upside down - to see it from a different perspective. In its most powerful form, inversion is asking how an endeavor could fail, and then being careful to avoid those pitfalls. [1]

With the advent of large language models (LLM), retrival augmented generation (RAG) has become a hot topic. However throught the past year of helping startups integrate LLMs into their stack I've noticed that the pattern of taking user queries, embedding them, and directly searching a vector store is effectively demoware.

What is RAG?

Retrival augmented generation (RAG) is a technique that uses an LLM to generate responses, but uses a search backend to augment the generation. In the past year using text embeddings with a vector databases has been the most popular approach I've seen being socialized.

RAG

Simple RAG that embedded the user query and makes a search.

So let's kick things off by examining what I like to call the 'Dumb' RAG Model—a basic setup that's more common than you'd think.